
#1790: [2019] Intel Studiosは複数人の同時パフォーマンスをボリュメトリック撮影できた
#1790: [2019] Intel Studios Could Volumetrically Capture Simultaneous Performances
![#1790: [2019] Intel Studiosは複数人の同時パフォーマンスをボリュメトリック撮影できた](http://voicesofvr.com/wp-content/uploads/2025/04/Voices-of-VR-Podcast.jpg)
#1790: [2019] Intel Studiosは複数人の同時パフォーマンスをボリュメトリック撮影できた
Voices of VR
要約
Voices of VRが2019年7月のSIGGRAPHで行った、Intel Studios事業責任者Sarah Vickへのインタビューの再公開回。冒頭でホストは、3D/4D Gaussian splatの台頭により大規模ボリュメトリック撮影ステージが復活し得るという見方を示す。Vickは10,000平方フィート・100台超のカメラを持つステージの狙い、ポイントクラウド出力とUnity連携、配信とビジネスモデルの課題、音声キャプチャ、スポーツ活用、倫理観までを語った。
この回のポイント
- ●ホストは、Intel Studiosのような大規模ボリュメトリックステージが、Gaussian splat(特に4D)の普及で再評価されると見ており、今年のVenice Immersive上位作品が3D/4D Gaussian splatを使っていた点を根拠に挙げている。
- ●Intel Studiosは約10,000平方フィート、100台超のカメラを備え、アバター単体ではなく複数人の相互作用を捉える大規模撮影を強みとしている。
- ●出力はIntel独自の圧縮形式によるポイントクラウドのボリュメトリック動画で、UnityやHoudini用プラグインから再生でき、Kira Benningの『Running』ではボクセル風の表現に加工された。
- ●Vickは最大の課題を配信(ディストリビューション)とし、案件ごとにモデルが異なる現状から、少数の提供モデルへ絞る必要があると述べている。
- ●音声面では、ゲームエンジン内でのダイナミックな再生や空間音響の研究を進めているが、ステージは映像用に設計されており録音環境の改善が課題だという。
- ●スポーツでは、技の分析や将来的なライブのボリュメトリック配信を見据えており、著作権や倫理など規制面の遅れにも議論が及んだ。
章立て
導入:大規模撮影の復活予想
ホストが、Intel Studiosの大規模ボリュメトリック撮影と、Gaussian splatによる再評価の可能性を紹介する。
Vickの経歴とIntel Studiosの成り立ち
物理学出身でイノベーション戦略コンサルを経た経歴と、スタジアム用ボリュメトリック技術を起点にした2018年1月のスタジオ立ち上げを語る。
『Running』とポイントクラウド/ボクセル表現
ポイントクラウド出力と独自圧縮Volaを使い、開発者がサイケデリックなボクセル表現に仕立てた経緯を説明する。
他方式との比較と大規模ステージの強み
DepthKitやMicrosoftの方式などとの違い、10,000平方フィートで100台超のカメラという規模の意義を述べる。
ストーリーテリングと配信・ビジネスモデル
没入型の物語手法は未確立であり、収益化と配信の経済圏づくりが急務だと議論する。
音声キャプチャと出力形式
空間音響の取り組み、ステージの音響上の制約、ポイントクラウドの納品方法を説明する。
スポーツ活用とライブ配信の展望
Intel Sportsとの連携、技術分析やライブのボリュメトリック配信、視聴者への制御権の移譲を論じる。
規制・倫理と没入技術の可能性
著作権や倫理の問題、個人の倫理観の役割、没入技術が共感を広げるという展望と今後の見通しを語る。
解説記事
Voices of VRの再公開回で、ホストのKent Bye氏が2019年7月のSIGGRAPHで行った、Intel Studios事業責任者Sarah Vick氏へのインタビューを取り上げた。冒頭でホストは、このインタビューを「アーカイブ」として振り返る意図を説明している。Intel Studiosはロサンゼルスにある大規模ボリュメトリック撮影施設で、同時に約20人を撮影できたが、データ量が膨大で確実に配信する手段がなかったという。
なぜ今あらためて振り返るのか
ホストは、Gaussian splat(多数の半透明の点状要素で3Dシーンを表現する手法)の登場により、ボリュメトリックに撮影した生データから、高解像度の「4D Gaussian splat」を出力できるようになったと述べる。Apple Vision Proのような機器で再生できる品質に達しうるという。さらに、今年のVenice Immersive上位3作品が3D/4D Gaussian splatを使っていたことを挙げ、多人数が同じステージで同時に演じる大規模撮影が復活するという見方を示している。
Intel Studiosの特徴
Vick氏によれば、Intel Studiosは2018年1月に始動した。Intelが出資したReplay Technologiesのボリュメトリック技術はもともとスタジアム向けで、世界で約24のスタジアムに設置され、主に放送の演出強化に使われていた。同社はこれをスポーツ以外にも広げ、人間のパフォーマンスをニュアンスごとデジタル領域に持ち込めると考えたという。
ステージは約10,000平方フィートで、100台超のカメラを備え、増強を続けている。Vick氏は、多くのボリュメトリックステージがアバター単体の撮影に注力する中、人と人の相互作用を捉える点が差別化だと語る。運用には多くのIntel製シリコンが必要で、この規模はIntelだからこそ実現できると述べている。一方で、唯一の解が全てを満たすとは考えておらず、DSLRとスマートフォンの比喩を用いて、用途により複数の手法が共存すると説明した。
出力形式と『Running』
出力は、独自圧縮形式「Vola」によるポイントクラウドのボリュメトリック動画で、UnityとHoudini向けのプラグインで再生できる。Kira Benning氏の『Running』では、開発者がポイントクラウドの各点をボクセル風のオブジェクトに置き換え、サイケデリックなダンスフロアを作った。Vick氏は、このデジタルな美的選択が、実写的な人物と非物理的な世界の落差、つまり不気味の谷のような違和感を和らげたと見ている。
配信とビジネスモデルという壁
Vick氏が最大の課題として挙げたのは配信である。同氏は、案件ごとに映画スタジオ、インディーアーティスト、ブランド、スポーツリーグなど異なる契約形態があり、いわば「8つのモデル」を扱っていると語る。作品がスポンサー付きなのか、プラットフォームで販売されるのか、LBE(ロケーションベース・エンターテインメント)で巡回するのかを常に問い、持続可能な経済圏が必要だと述べた。そのために、データのさらなる圧縮、パイプラインへの読み込みやすさ、低い計算要件を通じて、スタンドアロン機やスマートフォンでも再生できる方向を探っている。
ホストは、これを技術普及曲線の一段階として位置づけ、映画祭やLBE、博物館といった特注型の段階から、配信と経済の仕組みが整えば主流へ渡れるという見方を示している。
音声、スポーツ、そして倫理
音声面では、ゲームエンジン内でのダイナミックな再生や、環境による響きの違いを再現する空間音響の研究が進行中だという。ただしステージは映像のため光を均一に反射させる設計で、音の収録には最適でないと認めており、個別マイクとアンビソニックの併用を試している。
スポーツでは、姉妹組織のIntel Sportsと連携し、技のフォームを全方位から分析する用途や、将来的に選手の隣に立つような体験を構想している。ライブのボリュメトリック配信はTrueViewのチームが取り組む領域だと述べた。話題は、ファンに演出の制御権を渡す可能性や、二次創作と著作権、規制の遅れ、個々人が倫理観を持つ重要性にも広がった。
まとめ
この回は、当時の課題だった「大規模な撮影はできるが届ける手段がない」という状況を記録した証言であり、ホストはGaussian splatによって状況が変わる可能性を指摘する。日本の読者にとっては、撮影技術そのものより、圧縮、再生環境、収益モデルまで含めて整える必要があるという示唆が重要だろう。なお撮影ステージの現状や、Gaussian splatを使った具体的案件の詳細は、この回では語られていない。
文字起こし(英語・自動生成)
The Voices of VR Podcast. The reason why I'm pulling this out is because Sarah was at the Intel Studios volumetric capture studio, which was like this huge space in L.A. that could capture volumetrically like 20 people at a time. The problem was that there was a huge amount of data and wasn't really feasible to have any way of reliably delivering it. There were some pieces like running that was at Sundance of 2019. But I actually ran into a project that may have actually used this studio because now that there's Gaussian splats, There's new ways of basically taking this volumetrically captured data and taking that same raw data and being able to output something that's like a 4D Gaussian splat, which can have high resolution and be good enough quality to be able to run on like an Apple Vision Pro.
So I wanted to dig into this interview just as a bit of an archive, because I actually do think that there's going to be a comeback of this type of large-scale volumetric capture stage, especially considering the top three projects at Venice Immersive this year both had 3D Gaussian splats as well as 4D Gaussian splats. And so basically volumetrically captured performances on high resolution using the Gaussian splat sort of like algorithms and having lots of people perform at the same time on the same stage, I think is just something that's going to be very useful. And there's going to be a lot more people that are going to want to do something like that. So we're covering all that and more on today's episode, the Voices of VR podcast. So this interview with Sarah happened on Tuesday, January 30th, 2019 at SIGGRAPH in Los Angeles, California. So with that, let's go ahead and dive right in. My name is Sarah Vick. I am the head of business for Intel Studios. Intel Studios is a volumetric capture facility in L.A. and also an immersive media studio, and I run all of business development as well as some of our finances.
So maybe you could give me a bit more context as to your background and your journey into immersive technologies. Yeah, so my journey into immersive has been unexpected, serendipitous, fantastic. I actually started out in physics. My undergrad degree was at Stanford, and I worked in a laboratory doing ultra-low-tem physics, a really interesting field. And I knew I didn't want to live in the physics world just because, to be honest, I'm an extrovert and living in a sub-sub-basement. Turns out it's really lonely down there. So I quickly pivoted into business and went into strategy. So I was an innovation strategy consultant for four years in Boston, working with big firms, small firms, all launching new products, new technologies, trying to understand new industries. And that ended up kind of positioning me relatively well to move from business school into the immersive world because there's no place for digital disruption like immersive media right now.
I mean, this is upending an entire industrial complex, I should say, that's been established for nearly 100 years. And it's time to put that on set. Yeah, it's very interesting being here at SIGGRAPH because I really see all these worlds converging from the game engines, the live real-time engines and Unity and Epic and just this confluence of Hollywood and game developers, but also all the different volumetric capture solutions that are out there. There seems to be the 360 video to the motion capture to all the other different approaches to trying to capture volumetric objects. And it's kind of refreshing to be here at SIGGRAPH because it's like, oh, yeah, by the way, all these immersive technologies have always been used within Hollywood for all the visual aspects. But there's a whole industry and pipeline where what we would, and from the VR community call, like these immersive tech, is kind of embedded into the production of Hollywood and these entertainment,
but also those same tools being fed into how you create these real-time game engine experiences in immersive environments. And so I just see, like, more and more that a lot of this content is moving into more of a real-time paradigm. And so Intel has this very interesting approach of doing volumetric capture, and I know that Kira Benning had a piece at Sundance this year called Running that I was able to see and very much enjoyed and saw the real power of what you're able to do with having lots of people in a space at the same time moving in synchrony. So maybe you could set the context a little bit more in terms of the story of Intel and how it came about that you have this whole Intel Studios and this capture solution that you have. Yeah, absolutely. I think you hit on a ton of great points there. To be honest, there's this really interesting transition where we're moving from 2D creation, where the frame is my tool as a director to 3D content creation. And for a lot of Hollywood, it's already been in place for a long time, like you said, in kind of the VFX world.
And I think that the more and more it's becoming a tool set that we get to use on a day-to-day basis and not just for, like, the special effects and special scenes. So Intel Studios is about 18 months old at this point. It's July 2019. So we launched really in January of 2018. And Intel actually invested in a company called Replay Technologies that had volumetric capture, and we started doing that volumetric capture in the stadium. So there's actually nearly two dozen stadiums outfitted around the world with volumetric systems that capture that 3D data of the field, and then you're able to move through that scene. Most of that content is going to broadcast enhancement today. We knew that there was applications outside of sports for this technology and really believe in, frankly, being able to bring that human performance from the stage into a digital realm in order for someone to interact with it in 3D. And I think that's what makes volumetric so different is that being able to capture that full human performance and not just, you know, either a smoke cap skeleton, although you can get very detailed skeletons, and then but really getting the full nuance of the performance.
Yeah. Well, the thing that I noticed from Kira's piece of Runnin' is that she tried to create this feeling of being at a dance party and having people moving in synchrony. And so to be able to do a volumetric capture of music playing and people moving in synchrony and then to kind of replicate that and copy it in to make you really feel like you're in a dance party. Of all the different VR experiences I've had, that's probably the closest that I've felt to actually being at a dance party. But there's also, from what I understand, the fact that it was voxel-based volumetric capture allowed Kira to add all these really wild shaders that gave this whole other ethereal, just otherworldly surreal dreamlike quality to the experience. And so I'm just wondering if you could kind of break that down a little bit of that decision to use voxels and sort of the advantages and disadvantages of that, because I know there's a lot of different volumetric solutions and the voxel is a very specific design choice to go with.
So I'm just curious to hear a little bit more about the underlying voxel, the underlying decision to use voxels. Yeah, absolutely. I think, by the way, the technical term for the style chosen is trippy. But it's like definitely Reggie's influence coming in there too. Yeah, very psychedelic as well. Incredibly psychedelic. It was like every time I get in there, I'm just like, ooh, that was a nice little relief. So the point that we use is we actually use Point Cloud as our first output. And then from there, it's built into basically a Unity plug-in of using a proprietary compression format called Vola. and that developers can take and then manipulate however they want. In the case of Reggie, you're exactly right that they took each of those point clouds and then represented it as a secondary object. And the developer, Reza, who actually did that work, is just spectacularly talented in terms of what he was able to do. And I know I talked to a lot of you.
I'm not technical myself and I don't develop in Unity, but I talked to a lot of community developers who were really blown away by what he was able to achieve with it. And I think that for us, it is that question of how do you translate human performance into a digital scene? And in this case, it was very digital in its aesthetic choices as well, right? It really looks like a, it was kind of a disco floor, but very quadratic and square. And so it was no way kind of a more analog scene, which you could potentially go the alternative route. And so I think the interpretation of the point cloud into these kind of voxelized characters ended up helping bridge that gap for people from an aesthetic sense, that they understood what they were looking at, that they were looking at people dancing, but at the same time it wasn't that jarring, uncanny valley of, okay, this is like people dancing,
and they look real, but they're inserted into this totally psychedelic, no-physics world of Reggie Dancefloor and Kira Dancefloor, and I think that that kind of bridging between those two worlds is helpful for them. so there's a number of different volumetric capture solutions that are out there ranging from just regular motion capture and then be able to get like the almost the raw data of the motion and then it's kind of up to the individual to create either a volumetric rig with textures that you then you are able to do all sorts of variety of different things that you can turn people into superheroes and all this other amazing stuff that is possible but there's also like depth kit which I think is probably on the other extreme in terms of doing volumetric capture with just regular either RealSense cameras or with the Kinect cameras. And then you have, like, the Microsoft Mixed Reality capture where you have this ability to use a lot of SLR cameras and to kind of do one person at a time but do, like, the super high resolution.
And there's LiDAR scanning. There's all sorts of other approaches of AI extrapolation. and imagine that there's going to be continued lots of different solutions that are out there, but they all have their different advantages and disadvantages. And so how do you see that the Intel volumetric capture stage, how that relates to all these, a variety of other different volumetric capture solutions? Yeah, I think it's a great question. And to be honest, I don't think that we are, we don't adhere to a belief that there is one solution that's going to fit everything. Just like, to be honest, my brother has a beautiful DSLR camera that takes fantastic photos, and I have a Pixel 2. Like, those two, both of them are able to capture reality and bring them into my digital world and then preserve them or I'm able to manipulate them in different ways. And I think the same is going to exist for 3D data, right? Because fundamentally, what are we changing? We're changing the mode of capture from being 2D to 3D. And so I think there'll be a lot of solutions out there, and there'll be a lot of different solution finding when you kind of go into production,
that you'll go into a choice of, okay, am I using 16 mil, or am I using a RED camera, right? Filmmakers make these choices all the time too And there be similar decisions made along the production path in three capture as well In the case of Intel Studios we started with the belief that large scale is going to be really critical for capturing not only larger scenes, but also the interaction between people. Because so many of the volumetric stages were focused on avatar capture, and maybe you can fit two people into the Microsoft solution, but it's pretty tough. And even then, if you do, it's a very intimate scene, two people talking. And so for us, we wanted to give creators the ability to open up that stage. So the stage itself is 10,000 square feet. We have over 100 cameras on the rig today, and we're pushing that northward kind of every week. And I think for us, the scale was something that only Intel could do.
It takes a lot of Intel Silicon to run the volumetric stage, and so our resources are making us uniquely positioned to deliver that solution. But at the same time, we're constantly testing new ways of capturing within that stage. Now, we're maintaining the kind of we have a laboratory of capture, but we're expanding into audio capture. We're expanding into a little bit of, like, closed-in capture, getting a smaller stage. So we're experimenting with that all the time, and I think fundamentally our belief is that this ability to bring 3D creation to human and live performance is going to be the real tenant of it. It's interesting to hear that it seems like it's in this stage where there's still a lot of, like, the way I think about it is that there's these different phases of innovation that happens where there's technological innovation where the Intel volumetric capture studio is available.
Now I go to the creators, like, what can we do with this? And then listen to what the creators, maybe create a demo or prototype, like the Kira Bennings, the run-in, and then you're like, oh, this is what's even possible with this. And then you're like, oh, what can I do with a bunch of people now? And then it sort of opens up the possibility. And then usually there's also the distribution aspect of how you get this into people's hands, but then actually make it into a format that can fit onto either a mobile phone or if it's on a PC or if it's only in a museum context. And so that seems to be a part of this feedback loop of getting it to the point where it can kind of go from the PC and eventually get onto the more and more self-contained, tele-less mobile devices. And then finally, the audience being able to watch it and consume it and know how to experience these different pieces of content. But that feels like we're kind of still at that phase of, like, trying to figure out, okay, now that this exists, what can we do with it? Yeah, I think you actually bring up a really interesting question of does storytelling, does successful storytelling, or does successful distribution come first?
And to me, it really feels like this is one of the biggest questions we're facing in this space right now, is that I have, I mean, fundamentally, no one knows the exact right methods to do immersive storytelling. And everyone's experimenting. Everyone's pushing the balance. I've seen some spectacular pieces this year. I mean, the piece, I'm going to butcher the name, but it was at South by Southwest featuring the woman in the wheelchair. I think it's called Metro Vente. I don't speak better. I think it was like four foot tall in the English translation. It was about, you know, it was at Sundance as well where it was shot by, the camera was actually four feet, and then you see the perspective of her as someone who's in a wheelchair, yeah. Yeah, and that piece to me, I mean, to be honest, I had kind of written off 360 video. I thought I was like 360 video is trash. Like I was only fixed off, right? And that piece changed my perspective on 360 video because it was so believable in the fact that you were sitting there and, yeah, you didn't have six degrees of freedom because you're in a wheelchair.
And they used animation in this beautiful way to draw your attention and allow the user to experience time passing without wondering if you were still looking in the right direction, which I find is also really challenging in 360 to know that you're allowed to look where something is not happening. and I just thought that piece was spectacular. I think Gloomy Eyes is similarly a six-stop kind of exploratory piece where you can really go through every single nuance of the scene as a giant. But fundamentally, I don't think that we've solved the equation for storytelling. To be honest, filmmakers haven't solved the equation for storytelling. They're still pushing the bounds of their medium, but they're a lot further along than we are in the immersive filmmaking world. And so I think there's progress being made there at the same time as we're opening up these new distribution channels, right? When I, every single time that I'm bringing a new partner on, my number one question to them, usually they have a fantastic story concept, right?
Most of the time I'm speaking to these incredible creators. And then the next question I have is, okay, where does this live, right? And is your vision that it'll get sponsorship? Is your vision that someone's actually going to buy this for purchase on a platform? is the vision that it's going to tour as an LBE for the next year and a half. But fundamentally, we have to build an economy around this because it's the only way that it becomes sustainable so that I can keep creating fantastic pieces and, frankly, keep investing in something like Intel Studios. So I struggle with that. I'm definitely seeing a little bit of both, that we're trying to push them both down the runway at the same time. Well, I think that what I've seen at least is that it has been the film festivals for the last five years that have been this ground for experimentation for the artists. So the artists often will be working with the technology and just pushing the technology to its limits of what it can do with the stories that they want to tell. It's like opening up new possibilities to be able to tell stories in ways that have never been told before
and how to kind of figure out this larger storytelling language and how to actually construct these different experiences, this has been a part of what has been iterating on for the last, I'd say, for five or six years at Sundance and Tribeca and South by Southwest and Venice Film Festival and Doc Lab and all these other just artists and storytellers that are out there experimenting. And that we have kind of run into this situation with a lot of this work has been difficult to get into the hands of people because of all the different limitations of it costs like $1,000 to $2,000 for a VR headset with all the equipment, and they set it up, and so there's a lot of barriers to entry. And so with Oculus Quest, there's these new distribution opportunities, but it feels like that I see this as a normal technological diffusion curve with any technology. If you look at cell phones or anything, it's like usually Simon Wordley has a model where it goes to like these four phases where it's like you have an idea, a concept, and a duct tape prototype that, like at SIGGRAPH here,
we have a ton of duct tape prototypes that are being soldered and broken and, like, just-in-time kind of functionality where it's just working just in time for the actual conference. And then you have, like, the custom bespoke enterprise applications that go into, like, anything that gets created is very handcrafted and bespoke and in a very specific use case for the enterprise and military. And then eventually you have the consumers sort of crossing the chasm into the mainstream, and then you have this, like, consumer product that is productized, and then all the other, you know, economies and apps, all the app stores and economic incentives to be able to make it sustainable, and then eventually get to mass ubiquity. But we're kind of still in this phase where we have the consumer hardware. It's in people's hands, but it's not, it's still kind of in this customized, bespoke, location-based entertainment, museums, these film festival circuits, and, you know, still a lot of resources to be able to run these experiences, but where I see it's on a trajectory of kind of crossing that chasm into the mainstream
as soon as you kind of figure out the larger distribution and economic ecosystem to be able to really sustain it. Yeah, I think bespoke is a great word for what I'm experiencing right now. And frankly, that's how we're creating content, too. A lot of people come to me as kind of the head of business for Intel Studios and say, hey, what's the model? Like, what's the model? one I'm like no I have eight models last year I cut a deal with a major film studio with an indie artist with a brand and with a sports league I mean like you couldn't paint four more different pictures trust me doing a deal with an indie artist and one with the film studios are really really different um and I and I think that they all have use cases in this space but everyone's figuring out how do I interact in immersive media? What am I bringing to the table in this new medium? And I think that that's actually really one of the things that's so exciting about being in this space right now is that we're getting, and I've heard others on your podcast talk about this,
is getting to define the way that we operate, getting to define the way that this medium comes to light. And I think there's really exciting opportunities both around the storytelling aspect and around the business aspect. Yeah, the audio aspect is something that I haven't heard before and is very interesting because I'm really excited to see the future of, like, spatialized music, like what's it mean to compose spatialized. But in this case, you're kind of capturing live audio. Like, I imagine there's going to be a lot of really interesting stuff where you're capturing stems and music and be able to actually do compositions that are interactive and dynamic and that you're able to really spatialize it within the game engine. but this sounds like you're actually able to do some potentially really interesting things of doing this full ambisonic type of audio capture, capturing the sound field from lots of different perspectives. And so what do you expect that you're going to be able to do with that? Oh, I'm actually so excited to get to talk about this
because I think if this year was the year of kind of first look at volumetric as an experience from a visual standpoint, Next year, the integration of audio is going to totally change some of the pieces that we have out there. Two of the partners that I'm working with this year have audio being a major feature of what the experience will be. One is actually going to have dynamic play in the game engine, which I think will be brilliant because it's going to really drive the person to interact with the medium all the more when they realize that they have that control. It's such a rare thing to take control of. and the other one is actually looking more on kind of the spatial audio and actually mimicking the taking the ability what does it sound like for me to be in this hallway versus what does it sound like for me to be in my office versus on the beach right and understanding not only what is the asset sound like itself but then how are they responding to their local environment and we have some great partners
that's a little bit more on the R&D side frankly but some great partners who are really looking into that and understanding the localization question for audio, which is a huge, it's one way in which it's actually rare that we think about that from a light perspective, even though I am, you know, the way I am perceived is because of the amount of light that is bouncing off of all of the surfaces around me, but it's also mostly coming from sources. And so I think audio is a different equation there, that so much more of the audio experience is about bounce, is about the resonance of the space that you're in. Does the Intel Volumetric Capture Studio have soundproofing so that you trying to minimize all the second third and fourth order reflections of the area so you were able to get like some real almost like studio audio that then you can like reproject into different virtualized spaces Yeah, so we designed the stage first for visual capture, which means that we wanted a lot of bounce. And so we have panels.
If you've ever seen an image of it, you actually see the panels all around the dome, and that's in order to increase an even light bounce and reduce shadows for the actual capture. That, of course, does not then drive the very best in audio pickup. And so we've been experimenting with how do we then improve the stage and move to the next phase where we are able to capture that with a little bit more of a natural surrounding, or at least an isolated surrounding. Yeah. Yeah, because I imagine that, like, a big part of talking to different audio engineers is that, like, the next phase of being able to do these second and third and fourth and fifth order levels of reflections by designing virtual spaces, but then to be able to bounce that audio out. And so I would imagine that if you did want to have, like, the flexibility to be able to switch between different contexts, that you would want to have a certain amount of clarity in the audio source files, that then you could do that. And most of the time that I've heard that people are micing people individually and then sort of broadcasting that out, but because you have all the sort of ambient mics in an array,
which is really like this live ambisonic audio that I imagine that you're able to at least do some interesting things with it as long as the audio is clean enough and not getting too many, like, interference from all the different reflections. Yeah, I think that's exactly right. And we do do both, and we mix them, and we're experimenting with what's the best. And that's really, I think, that kind of what I've loved about working with this team. And so for context, I have a tiny team, right? It's like a 20-person operation. But we're pushing the balance every day about how do we do the very best immersive capture that we can, whether that's visual and audio or who knows. Someday I'm really excited to get to do texture capture of some kind when we invent that and run it on Intel Silicon. So I think that it's all kind of an open for exploration at this point, and I know that there's a lot of people who are super, super smart that we're leveraging on the audio side. So maybe you could talk a bit more about the actual output. You know, you're getting all these, what I presume are digital SLR cameras.
You're taking photos and then, like, doing some processing. And then what are you actually delivering to the creators? Yeah, so Intel Studios produces point cloud. So we are producing basically volumetric video of individual frames of a compressed point cloud. Like I said, we use our own compression technique, which was Intel developed. It actually is one of the areas where we adopted something out of the labs from broader intel and brought it into studios, which has been wonderful to get to leverage that research base. And then we have a Unbi and Houdini plug-in that plays back that volumetric video. So most of the time we have some in-house development, but we also have a lot of partner developers who then take both the plug-in and then the volumetric video sequences and are able to play them back natively inside of their game engine. Great. And so what are you seeing here at SIGGRAPH? And what kind of, like, either conversations or interactions are you having here?
I think, so this is my first time at SIGGRAPH. So I'm excited to get to witness this madness. And I do think it's that intersection of gaming and Hollywood and technology that's so intriguing and causes interesting questions, right? Because there's a lot of theories about how the film is going to evolve to become more interactive. But I'm of the opinion that, like, books didn't go away just because films came about. We're going to keep films forever. But there is this new opportunity to create a new storytelling method where there's a lot more agency involved. And what does it even look like? Should we call it film? Should we call it, will actually games just evolve to be something where there's more narrative on the gaming side? And I think there's a lot of questions that are still unanswered. And having that dialogue with people and understanding where, particularly on the gaming side, where their heads are at,
I'm really intrigued to hear more about it. And so for you, what are some of the either biggest open questions that you're trying to answer or biggest problems you're trying to solve? Yeah, I'll be frank. My problem is distribution today. I've been probably mulling on, chewing on that net for a couple of months now and trying to figure out how do we get to a more turnkey solution to go from bespoke to, you know, maybe only five or six models of delivery. And I think that there's a lot of different options out there. One would be to just create more turnkey solutions for LBEs. I think there's more opportunities, too, about getting data, more compression on data, easier to load into pipeline, lower computational requirements on these games or on experiences in general so that people can load them onto a standalone
or onto their phone if they so desired, which not many people do. So I think that that's something that I've been thinking about a lot with my team and talking to partners about and trying to test new models about, okay, is this something that we could scale and operate on a global market? So that's been the main challenge that I've been tackling in the past, like I'll say working three to six months timeline. So, yeah. So you mentioned that you have a sports team. Is there any additional context that you could talk about, like, what the sports team would want to do? Is this for, like, training or research or advertising, marketing, or what type of things, another context and situations and industry verticals are you starting to tap into there? Yeah, so I'm lucky enough to have Intel Sports as kind of my sister organization, and so they are partnered with the NFL, with the NBA, with La Liga, with the EPL, with a huge array of sports teams and leagues.
The partner that I'm not allowed to disclose quite yet who the partner is, but it's a soccer team. And we're really, really excited because we think that there's hunger from the consumer base to actually engage in content and understand how to do tricks and skills. And I think immersive media has a great opportunity to actually dive into, okay, how do you do that? That triple axel, backflip, jump, whatever that Simone Biles just did on NBC the other night. I want to see that. I want to analyze it from every angle and understand what was the technique that she had for turning at that rate. So I think that sports sees a lot of opportunity in that world from the studio side where they have to capture. They're using it as a laboratory as much as we are. right? In the stadium side of content, there's crazy opportunities, I think slightly longer down the road, but we're starting to experiment with this too, where you can get in the game, literally,
like put a headset on and go stand next to Tom Brady. And I think that the vision for that is pretty clear. The technology will be ramping in the next five years. Yeah, when I watch different sports experiences, I'll see like the whole Intel, RealSense, like doing a whole 360, you know, view of something and kind of spin around and you get to see a replay. It's kind of a neat novelty. But I can imagine that with all the advent of artificial intelligence and AI and the fusion with all this capture that we're going to eventually get to the point of potentially having like a real-time live stream of a volumetric, get into an experience. And, you know, there's obviously NextVR and other people that have been focusing on the live broadcasting of that. But I imagine that there's going to be a potential desire for people who are huge fans of these types of experiences to want to actually be immersed in a volumetric capture. And I think the key there is that it kind of needs to be live in certain ways because it's like it's happening right now.
And there's already a really great experience of watching it on a 2D screen. And I think it would need to kind of at least match that of, like, feeling like it's happening right now and it's emergent and it's, like, happening and you want to sort of participate in that moment. And it would kind of have to get to that point. But imagine with all the different AI and deep learning and computer vision stuff that's being fused here at stuff like SIGGRAPH and what I'm seeing here on the floor and just in general in the larger community, that it seems like a technological inevitability that we'll get there at some point. I think so. I mean, Moore's Law is a beautiful thing. Intel is in this business for a reason, and part of it is that we do know that we'll be able to accelerate workloads like this. And the vision and milestone of streaming volumetric video is absolutely there on the Intel support side, the technology called TrueView, and that team has their work cut out for them. I feel somewhat lucky to get to work on a little bit more of a controlled stage while they work in NFL arenas.
but that I'm excited for fans and what they'll get to see. I think there's an interesting concept, too, of where do we end up handing control to the user, right? In sports, this is a particularly interesting statement because the broadcasters have had control of the camera for so long, right? And they're really, really, really good at what they do. NBC, ESPN, Turner, they're all fantastic. at producing a sports game the way that we are traditionally used to consuming it. So what's going to be interesting is with kind of the iGen coming up, who's more used to controlling content as they go, then is there an opportunity to hand some of that directorial, that producing skill, over to someone who's been publishing on Musical.ly for the past six years, right? Or maybe not that long. And I think it's still yet to be seen how much that generation comes to engage with content creation in this space.
But there will certainly be the opportunity. Yeah, earlier today I watched Tim Sweeney from Epic talk about the future vision that he has in the metaverse and to just hear everything that's happening with Fortnite. And this is the generation that has grown up with Minecraft. and spent, you know, for 10 years now, Minecraft has been out there and then moving on to Fortnite. We have this whole, like, social media and the youth and generation that's coming up, the Z generation or whatever they're being called, and they're immersed and engaged in media in a way that, you know, they just have this fluency for these virtual worlds and be able to interact with them. And so they are going to want to be in that world, be engaged, immersed into this level of control like a video game like Fortnite, and what is the future of entertainment going to look like. And I feel like to listen to Sam Sweeney talk about his vision of the future of the metaverse, I really take to heart a lot of what he's trying to create with the open standards, the interoperability, and an economy that is able to create this self-sustaining,
reinforcing aspect of those influencers of being able to give creator codes out and to have own sort of economics. So the depth in which there a whole economic consideration on top of the whole ecosystem of these virtual worlds just feels like we kind of moving into this whole new realm of what Tim Sweeney is saying We kind of on this roadmap towards this metaverse And so, yeah, just thinking about, like, how the entire media ecosystem is going to need to change from a lot of the way things are now and this kind of disruption of these immersive technologies. Yeah, and actually one of the areas that's most fascinating about this and someone who's, like, not ready for this at all is the regulatory bodies. right and think about copyright and the way that copyright works that nobody is enforcing copyright with any of the little you know 17 year old 12 year old creators and if you did you you just waste the amount of legal resources from an i mean yeah so it's it's a serious problem
that i feel like the industry like kind of the the status quo industry players are not willing and ready to face down what the new regulation is going to look like because they're holding on too close to the old regulation. And one of my favorite examples is, you've seen Rap Battles? I don't know if I have, no. Tell me about, what are those? So Epic Rap Battles, it started off as a YouTube series. Oh, yeah, I've seen those, yeah. And there's one, so we actually have an intern, Cam, who's phenomenal, and he's doing epic wrap bottles for cartoon characters. If you haven't seen it, go watch it. And really funny. It's like Sonic the Hedgehog and Mario bottling it out. Now, I'm a Sonic fan, but the question is, does Sega come after Cam for publishing that? Because they could, arguably. Please don't come after Cam.
But also, it's creating more buzz for them. It's creating this, you know, earned media is a term now. And I think that none of copyright law actually manages to wrap their heads around how that creator world will face. And I think there's going to need to be some serious changes. YouTube is basically skirting the issue. But we're going to have the same issue on all of the publishing platforms, Switch, Unity. Whoever ends up being the new broadcast method for a lot of these 3D mediums is going to face the same challenges. Yeah, there's certainly a lot of ethical frameworks and moral dilemmas that are coming up in this space, and I feel like this field is kind of the center of a lot of these larger moral dilemmas that happen with all of technology. So I feel like the regulatory bodies are a good 5 to 10, 15 years behind for where the technology is at, and so how do you kind of create a situation that is protecting the people that need to be protected
but also, you know, I think there's also a whole other level of, like, the larger sort of moral center of both the companies and the people that also, you know, may be kind of independently happening that is difficult to have a regulatory body. So I tend to look to Lawrence Plastic to see where he's looking at that there's, like, these four major, you know, vectors, that there's an economic vector, there's a cultural vector, a legal vector, as well as the technology and the code. And so how all those combine together to shape behavior and culture is, I think, something that has to kind of take this holistic approach. But, yeah, there's certainly no lack of moral dilemmas that we're facing with mixed reality. I have to say that there's a sense that we're more and more disembodying the moral right of companies. Companies are groups of people. Like, you are a person in this industry. You should be acting with strong ethics. Would your mother agree with what you're doing right now?
If not, then take a second glance. And I think that more and more we're disembodying these things. Oh, no, Facebook is doing wrong. Google is doing wrong. Well, you know what? I know a lot of people who work at both of those companies. And I have to ask them, do you believe in what you're doing as part of this organization? And do you believe that you're acting with the moral rectitude that we should all bring to every day? so I'm pushing for us to each embody our own ethical belief and position in this industry yeah that's a good approach I think that could also get dangerous if it's not like we're if you're not on the same agreement of like you know so the sort of moral the moral relativism of like sort of like you know if you if you value so for example if you value autonomy above all else then that should could have like you could be doing harm so I feel like it's actually like a need to take some sort of pluralistic approach to find out how to address these issues, either through ethical frameworks and guidelines, because there's different open philosophical
problems that have a multitude of very different perspectives. But what I find interesting about this emerging technology space is that you actually have to kind of collapse that infinite potential of that wave function into something that is an actual design decision. And I feel like that is where the rubber meets the road in that you have this undoubted potential of all these different possibilities, but yet you still have to set what your values are and what you're sort of trying to embody in terms of the moral virtues or the problems you're trying to solve. And that with each of those different contexts, you have different situations that have to be navigated in their own way. And so how do you take into all that input and also as the designers of people in this space, how do you take into consideration the multitude of different contexts and perspectives to be able to understand when you are trying to, like, make an actual decision and to build something that is functional and real and pragmatic to take into consideration the whole variety of the many different perspectives? Yeah, I think the collapsing function metaphor works well.
you know I am made up of a series of quantum particles all of those have a you know probabilistic function of their current location and time and yet if I punch you you are going to know that that hand was like in a very specific location and place and time so the I think that there's a question to me around you know how much of the frameworks will fall to individuals and how much to bodies, institutions that we can raise. Right now, to me, it feels very individual-based, just because the number of players in the space is still relatively small. So maybe that's why I fall more to that side right now. Great. And finally, what do you think the ultimate potential of immersive technologies are and what they might be able to enable? so uh this is a little bit more meta than uh i think the rest of my job feels very tangible but my belief in this space is is very wholehearted and fundamentally i believe that we are
expanding the ways that we tell stories and create experiences and i have a strong belief that if we were able to augment one's experience, we'd be able to become better people because I am fundamentally a computational machine. I make my decisions based on the experience that lives in my brain. And I grew up in a very rural part of Connecticut that's pretty homogeneous in terms of the experience that and I've benefited so much from watching indie film and from listening to other people tell their stories about where they grew up and understanding that not everybody else, Not everybody grows up in bucolic Connecticut. And I think that this represents another opportunity to expand that storytelling base, to create a world in which we are fundamentally nicer to one another. As silly as that sounds, I think that it would be nice to create a world where we are slightly nicer to each other.
I don't think it's silly at all because it's not happening. So if it was simple, it would be happening. That's true. And it seems at this moment in time as if we're on a trajectory heading the other direction, maybe because it's more simplistic and requires less technology. So I'm all for expanding our experience set and telling stories like they've never been told before. Great. Is there anything else that's left unsaid that you'd like to say to the immersive community? Keep on, keep it on. I think that I've been in this space now for just over two years, and I know that about halfway in that process, there was a period when folks really looked around, and it's been tough, right? We're all struggling to make business models work and make ends meet and get projects done, but what's incredible is that we still are, and we still will be able to. and I don't have any question in my mind as to whether or not this space is going to be successful and impactful.
And so if you ever need a pep talk, give me a call. Yeah, to me it's not a matter of if but when. And so I don't know when it's going to sort of get to that point, but I feel like when I come to the conferences like Seagraph, I see all the different larger ecosystem of the entire entertainment industry, the gaming industry, and just where this whole movement towards immersive entertainment is headed, I feel like it's sort of an inevitability. It's just, I don't know when in larger context and all the pieces that need to be in place to have the full kind of crossing the chasm in the mainstream and kind of the straw or the tipping point that really makes it go from where it's at to really kind of exploding exponential in that way. So I agree that I feel like it is going to happen. I just don't know when. I say 2025 or 2045, and come back and ask me at the end of 2020, and I'll let you know. So whether we're on a five-year scale or if we're going to sort of take a whole winter and kind of go to 2045. But I feel like that's a reasonable expectation that it's either going to continue to kind of exponentially grow until 2025,
or we're going to sort of hit this other point sometime in the future. Yeah. Mark Zuckerberg said, what, 10 years? So 2025, you're on the mark timeframe? No, I'm on my own, sort of just talking to a lot of people, yeah. Yeah, I think I like 2025. Let's go with that. Okay, we'll check in again and see how it's going. So, anyway, I just want to thank you for sitting down and joining me today on the podcast and to let me know what's happening with the Intel studios and all the future of what's happening with volumetric capture and all the challenges and, yeah, just get a state of what's happening in VR for you and Intel. So, thank you. Absolutely, and thank you. And thank you for doing this podcast. It's a favorite of mine, and I think it really adds value to the community. So thanks for that. Awesome. Thank you. Thanks so much for listening to the Voices of VR podcast. And I just want to say that I'm really trying to live into the transformation economy, meaning I want to give away all this information for free, make it public accessible for not only you but the entire community.
And if you find value in that, then please do consider supporting me on Patreon. You know, I'm on this journey of trying to reach my 1,000 true fans. And if I have just $5 a month from that 1,000, I'll be in a much more solid and stable financial position. So really need your help to be able to get to that point. So if you enjoy what I'm doing here at the Voices of VR podcast, then please do consider supporting me at patreon.com slash Voices of VR. Thanks for listening. Thank you.
番組の概要欄(原文)
I interviewed Sarah Vick about Intel Studios Volumetric Capture Stage 2018 - 2020 at SIGGRAPH in Los Angeles, CA on Tuesday, July 30, 2019. This is a listener-supported podcast through the Voices of VR Patreon. Music: Fatality
出典・この記事について
- 番組
- Voices of VR(Kent Bye)
- 元の回
- #1790: [2019] Intel Studios Could Volumetrically Capture Simultaneous Performances
- 配信日
- 要約作成日
- 公式サイト
- voicesofvr.com
この要約・解説記事は、元の音声を文字起こしした内容をもとに AI(Claude)が日本語で作成したものです。 固有名詞や数値に誤りを含む場合があるため、正確な内容は元の音声でご確認ください。制作方法について